OF ARTS AND SCIENCES. 251 



this way, the lines of Nobert's 1 9th band are about x^^^ott ^^ ^^ iii*^^^ 

 in width. The photographs made by Dr. Woodward seem to give a 

 little greater value. The finest lines I have succeeded in ruling are 

 about xB'oVoo ^^ ^^ i^ch in width. These values are substantially the 

 same as tliose given by Dr. Royston Pigott, as representing the ulti- 

 mate limit of visibility under the microscope. The smallest angle at 

 which an object can be distinctly seen is stated by him to be 6", while 

 other writers place it as high as 60", or even 120". Even the smallest 

 value named is much too large. I will at any time undertake to rule 

 a single line, ^qooT) *^f ^^ "^^^^ i'l breadth, which can be seen at the 

 distance of seven inches from the eye. This corresponds to an angle 

 of about 1''. In this case the line is filled with plumbago, but, if it is 

 reflected from a silvered surface, it can be easily seen at the distance 

 of eleven inches from the eye. Comparing minute particles of matter 

 which can be seen under a Tolles -jV objective with those which can 

 be measured, in the way indicated above, there is every reason to sup- 

 jwse that the limit of visibility falls beyond 4^oVo o ^^ ^'^ inch. It is 

 quite possible that the conclusion reached by Sorby, that the microscope 

 has already reached the limit of its power in separating lines whose 

 distance apart is equal to one half of a wave length, may be found to 

 be justified by future observations. It is certain that no lines heyond 

 Nobert's 19th band have ever been resolved. The great difhculty in 

 distinguishing true from spui'ious lines has caused more than one skil- 

 ful microscopist to doubt whether the resolution has been certainly 

 carried a far as that point. But that light is "of too coarse a nature " 

 to enable us to see particles of matter as small as ^-q^Vo^ ^^ ^'^ inch, is 

 a conclusion which can be refuted without the slightest difficulty. 



How are Noherfs Finest Lines produced'? 



In trying to answer this question, I shall give the results of four 

 distinct lines of investigation. Neither of these furnish conclusive 

 evidence, but they are all suggestive of possibilities. 



I. I have already stated that there is strong evidence that they are 

 ruled with a diamond having a knife-edge. To this is added a fact 

 derived from my own experience, and confirmed by a trial of several 

 months ; viz., that when a diamond, having a polished knife-edge, is set 

 slightly inclined to the direction of the lines rided, its riding qualities 

 improve with use. The diamond with which bands of 50,000 lines to the 

 inch were first successfully ruled would at first barely rule 10,000. It 



